Conference Agenda

The implementation of MEI support in MuseScore 4.2 marks a significant milestone in bridging the gap between the Music Encoding Initiative (MEI) and a widely-used music notation software application. While existing MEI creation tools exist, direct export solutions from widely-used notation software, such as MuseScore, have been lacking. This paper details the collaborative effort between the Digital and Cognitive Musicology Lab (DCML) at EPFL and the RISM Digital Center to integrate MEI support into MuseScore.

The implementation focuses on round-trip lossless conversion, enabling users to export and import MEI files seamlessly, facilitating data reuse and versatile use-case scenarios. It is based on MEI-Basic, a subset of MEI specifications, which ensured a well-defined starting point and coverage of the most essential features of Common Western music notation. While some limitations exist due to application design and differences in data structures, the implementation offers a foundation for further improvements.

However, challenges persist, such as metadata limitations and handling of IDs, which generate further tasks for future refinement. The collaborative effort acknowledges the ongoing role of the MEI community in providing feedback and insights to shape recommended workflows and best practices. The experiences gained by users employing MuseScore with MEI support will contribute to the continuous improvement and adjustment of this integration, fostering a more robust MEI ecosystem.

Utilizing the ElasticSearch library and the Java Language, researchers are efficiently and robustly parsing MEI in a fresh and flexible way, vastly improving on the previous working solution presented at the 2023 MEC by David Day from Brigham Young University. Decreasing indexing time, providing infinitely more options for searching, improving accuracy, and eliminating the need for constant re-indexing are some of the impressive improvements made. Our act of reducing MEI to a skeletal frame for rapid and variable searching has exciting implications for the community at large, and will become a crucial touch stone for those working on similar projects in the future.

The DFG project Digital Liszt Catalogue Raisonné, which was launched in 2020, catalogues and systematizes Franz Liszt's oeuvre. To this end, philological and bibliographic information on Liszt-related sources and works will be collected and encoded and stored in a suitable data format as the basis for the index. MEI, an XML-based format that is human- and machine-readable and equally suitable for long-term archiving and data export via API, will be used for this purpose. The catalogue raisonné to be developed will provide a selection of possible views of this data in a web interface. When indexing the source holdings, problems arise in the FRBR categorization, which forms the basis of MEI. The poster presents a problematic source in this respect and presents a preliminary coding in MEI for discussion.

The Liszt-related source situation is characterized by the fact that, in addition to finished versions, early or transitory states of the works are also well documented by a rich stock of sources. However, Liszt tended to use the same source materials for multiple revisions of his works. There are often years or decades between the individual revisions. In such cases, the various stages of editing the sources each form their own versions of a work, and occasionally a new work results from the editing of a source. For this purpose, Alexander Wilhelm Gottschalg's copy of a passage newly composed by Liszt is firmly attached to the print. A later, also published version was created by making further deletions and additions, deleting Gottschalg's copy and adding autograph leaves. The state of preservation described above makes intuitive recording according to the FRBR model difficult, firstly due to the assignment of the source object to different versions. A second difficulty results from the materially complex composition of the source consisting of printed and manuscript leaves.

The hybrid music edition project Reger-Werkausgabe (RWA) has developed new research and publication methods since its beginning in 2008. With its online platform “RWA Online” it has constituted a sustainable foundation for complex digital research work like the research field of “templates” that Reger used in different ways for his compositions. RWA concentrates on templates from literature und music in Reger’s Organ Works (module I), Songs and Chorale Works (module II) and Reger’s Editions of Works by other composers (module III).

Initially, RWA’s work about templates was designed to fit into the condensed spaces of a printed volume’s critical commentary. The unlimited space in the project’s digital publication part soon led to new presentation ideas that could serve to answer research questions like text comparisons or to allow for more comprehensive research about text origins. RWA ahs is compiled a thoroughly worked out data set that makes different levels of relations between Reger’s works and their template works accessible.

RWA had to solve several (technical) problems. Beside digitalization of template source material, a data model had to be. In addition, software solutions that make these contents accessible and usable from varying research interests or entry points had to be drafted, evaluated, and developed.

For its data model RWA facilitates standards like TEI, MEI and FRBR to document characteristics and relations. This data is than brought to live with the help of an eXist-db and four eXist-db web applications developed within the RWA that provide data access, data, text comparison as well as object visualization.

This paper will give an overview of the data model and software developments in the RWA by presenting key problems and respective solutions form its edition work. There will be furthermore insights into extensions that evolve from the work on the third project module.